使用 gganimate 时修复 blurry/overlapping 文本

Question

我在使用 gganimate 时无法创建高分辨率文件。图表中的文本变得模糊，几乎就像是为 .gif 中的每个新帧编写文本一样。

这个问题有简单的解决方法吗？我尝试摆弄 ani.options()，但也没有任何运气。

示例数据和代码：

value <- sample(0:30, 300, replace = TRUE)
year <- rep(c(2000, 2001, 2002), 100)

df <- as.data.frame(cbind(value, year))

library(gganimate)
library(tidyverse)

df <- df %>% group_by(year) %>% mutate(avg.value = mean(value))    

p <- df %>% 
   ggplot(aes(x = value)) + 
   geom_histogram(position = 'identity', color = "white") + 
   geom_segment(aes(xend = avg.value, x = avg.value, y = 0, yend = -.5), 
   color = "red", size = 2.5) +
   transition_states(as.factor(year), transition_length = 2, state_length 
   = 1, wrap = TRUE) + 
   annotate("text", x = df$avg.value, y = -1, label = "Average") + 
   labs(x = " ", 
   y = "", 
   title = "Example chart", 
   subtitle = "Don't it look blurry?")

animate(p, nframes = 100, renderer = gifski_renderer("example.gif"))

Answer 1

要向 annotate（或我的代码中的 geom_text）提交多个相同的 avg.value 值。当每年你只需要一个时，你就会添加一百个值。这导致一百 "Average" 相互绘制。
解决方案是向 geom_text（即 df2）提交不同的数据，这些数据仅包含 3 行（每年一个条目）。

代码：

library(gganimate)
library(tidyverse)
value <- sample(0:30, 300, replace = TRUE)
year <- rep(c(2000, 2001, 2002), 100)

df1 <- as.data.frame(cbind(value, year))
df1 <- df1 %>% group_by(year) %>% mutate(avg.value = mean(value))    
df2 <- df1 %>% group_by(year) %>% summarize(avg.value = head(avg.value, 1))    

r1 <- ggplot(df1) + 
   geom_histogram(aes(value), position = "identity", color = "white") + 
   geom_segment(aes(x = avg.value, xend = avg.value, y = 0, yend = -0.5), 
                color = "red", size = 2.5) +
   geom_text(aes(avg.value, -1, label = "Average"), df2) + 
   transition_states(as.factor(year), 2, 1) + 
   labs(title = "Example chart", 
        subtitle = "No more blur",
        x = NULL, 
        y = NULL)
animate(r1, nframes = 100, renderer = gifski_renderer("example.gif"))

使用 gganimate 时修复 blurry/overlapping 文本

Fix blurry/overlapping text when using gganimate

r

ggplot2

gganimate